# Linear Models for the Prediction of Animal Breeding Values, 3rd Edition.
# Raphael Mrode
# Example 11.2 p183
간단한 설명은 다음 포스팅을 참고한다.
Mixed Linear Model for Computing SNP Effects(SNP-BLUP Model), unweighted analysis
# Linear Models for the Prediction of Animal Breeding Values, 3rd Edition. # Raphael Mrode # Example 11.2 p183 SNP effect를 임의 효과로 다루어 SNP effect를 추정한다. 각 SNP effect의..
bhpark.tistory.com
Data
13 0 0 1 558 9 0.00179211 1.3571429 -0.3571429 0.2857143 0.2857143 -0.2857143 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857
14 0 0 1 722 13.4 0.00138504 0.3571429 -0.3571429 -0.7142857 -0.7142857 -0.2857143 0.7857143 -0.1428571 0.07142857 -0.1428571 -0.7857143
15 13 4 1 300 12.7 0.00333333 0.3571429 0.6428571 1.2857143 0.2857143 0.7142857 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857
16 15 2 1 73 15.4 0.01369863 -0.6428571 -0.3571429 1.2857143 0.2857143 -0.2857143 -0.2142857 -0.1428571 0.07142857 0.8571429 0.2142857
17 15 5 1 52 5.9 0.01923077 -0.6428571 0.6428571 0.2857143 1.2857143 -0.2857143 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857
18 14 6 1 87 7.7 0.01149425 0.3571429 0.6428571 -0.7142857 0.2857143 -0.2857143 0.7857143 -0.1428571 0.07142857 0.8571429 0.2142857
19 14 9 1 64 10.2 0.01562500 -0.6428571 -0.3571429 0.2857143 0.2857143 -0.2857143 0.7857143 -0.1428571 0.07142857 0.8571429 -0.7857143
20 14 9 1 103 4.8 0.00970874 -0.6428571 0.6428571 0.2857143 -0.7142857 -0.2857143 -0.2142857 -0.1428571 0.07142857 0.8571429 -0.7857143
1 ~ 3 : animal, sire, dam
4 : general mean
5 : EDC(using weight)
6 : Fat DYD
7 : EDC 역수
8 - 17 : SNP1 ~ SNP10의 coding하고 평균을 0으로 scaling한 값
(7 - 17 컬럼은 원래의 자료에서 계산을 하여 입력하여야 한다.)
* 계산 방법은 위 포스팅을 참고
Renumf90 Parameter File
# Parameter file for program renf90; it is translated to parameter
# file for BLUPF90 family programs.
DATAFILE
snp_data2.txt
TRAITS
6
FIELDS_PASSED TO OUTPUT
WEIGHT(S)
7
RESIDUAL_VARIANCE
245
EFFECT
4 cross alpha
EFFECT
4 cross alpha
RANDOM
diagonal
RANDOM_REGRESSION
data
RR_POSITION
8 9 10 11 12 13 14 15 16 17
(CO)VARIANCES
9.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 9.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 9.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 9.96 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 9.96 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 9.96 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 9.96 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 9.96 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 9.96 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 9.96
OPTION solv_method FSPAK
가중치로 7열을 주었다.
나머지는 설명은 다음 포스팅 참조
blupf90으로 Mixed Linear Model(SNP-BLUP Model) for Computing SNP Effects(unweighted analysis) 풀기
# Linear Models for the Prediction of Animal Breeding Values, 3rd Edition. # Raphael Mrode # Example 11.2 p183 간단한 설명은 다음 포스팅을 참고한다. 2020/12/19 - [Animal Breeding/R f..
bhpark.tistory.com
Renumf90 실행 화면
renumf90 실행 로그
RENUMF90 version 1.145
renumf90_snpblup_snp_w.par
datafile:snp_data2.txt
traits: 6
R
245.0
Processing effect 1 of type cross
item_kind=alpha
Processing effect 2 of type cross
item_kind=alpha
Reading (CO)VARIANCES: 10 x 10
Maximum size of character fields: 20
Maximum size of record (max_string_readline): 800
Maximum number of fields for input file (max_field_readline): 100
Pedigree search method (ped_search): convention
Order of pedigree animals (animal_order): default
Order of UPG (upg_order): default
Missing observation code (missing): 0
hash tables for effects set up
first 3 lines of the data file (up to 70 characters)
13 0 0 1 558 9 0.00179211 1.3571429 -0.3571429 0.2857143 0.2857143 -0.
14 0 0 1 722 13.4 0.00138504 0.3571429 -0.3571429 -0.7142857 -0.714285
15 13 4 1 300 12.7 0.00333333 0.3571429 0.6428571 1.2857143 0.2857143
read 8 records
table with 1 elements sorted
added count
Effect group 1 of column 1 with 1 levels
table expanded from 10000 to 10000 records
table with 1 elements sorted
added count
Effect group 2 of column 1 with 1 levels
table expanded from 10000 to 10000 records
wrote statistics in file "renf90.tables"
Basic statistics for input data (missing value code is '0')
Pos Min Max Mean SD N
6 4.8000 15.400 9.8875 3.7434 8
random effect 2
type:diag
Wrote parameter file "renf90.par"
Wrote renumbered data "renf90.dat" 8 records
renumf90 실행 결과로 생성된 파일
renf90.dat
9 0.00179211 1.3571429 -0.3571429 0.2857143 0.2857143 -0.2857143 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857 1 1
13.4 0.00138504 0.3571429 -0.3571429 -0.7142857 -0.7142857 -0.2857143 0.7857143 -0.1428571 0.07142857 -0.1428571 -0.7857143 1 1
12.7 0.00333333 0.3571429 0.6428571 1.2857143 0.2857143 0.7142857 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857 1 1
15.4 0.01369863 -0.6428571 -0.3571429 1.2857143 0.2857143 -0.2857143 -0.2142857 -0.1428571 0.07142857 0.8571429 0.2142857 1 1
5.9 0.01923077 -0.6428571 0.6428571 0.2857143 1.2857143 -0.2857143 -1.2142857 -0.1428571 0.07142857 -0.1428571 1.2142857 1 1
7.7 0.01149425 0.3571429 0.6428571 -0.7142857 0.2857143 -0.2857143 0.7857143 -0.1428571 0.07142857 0.8571429 0.2142857 1 1
10.2 0.01562500 -0.6428571 -0.3571429 0.2857143 0.2857143 -0.2857143 0.7857143 -0.1428571 0.07142857 0.8571429 -0.7857143 1 1
4.8 0.00970874 -0.6428571 0.6428571 0.2857143 -0.7142857 -0.2857143 -0.2142857 -0.1428571 0.07142857 0.8571429 -0.7857143 1 1
renf90.par
# BLUPF90 parameter file created by RENUMF90
DATAFILE
renf90.dat
NUMBER_OF_TRAITS
1
NUMBER_OF_EFFECTS
11
OBSERVATION(S)
1
WEIGHT(S)
2
EFFECTS: POSITIONS_IN_DATAFILE NUMBER_OF_LEVELS TYPE_OF_EFFECT[EFFECT NESTED]
13 1 cross
3 1 cov 14
4 1 cov 14
5 1 cov 14
6 1 cov 14
7 1 cov 14
8 1 cov 14
9 1 cov 14
10 1 cov 14
11 1 cov 14
12 1 cov 14
RANDOM_RESIDUAL VALUES
245.00
RANDOM_GROUP
2 3 4 5 6 7 8 9 10 11
RANDOM_TYPE
diagonal
FILE
(CO)VARIANCES
9.9600 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000
0.0000 9.9600 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000
0.0000 0.0000 9.9600 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 9.9600 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 9.9600 0.0000 0.0000
0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 9.9600 0.0000
0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 9.9600
0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
9.9600 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 9.9600 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 9.9600
OPTION solv_method FSPAK
가중치 열이 2열로 재배치.
위에서 생성된 renf90.dat와 renf90.par를 이용하여 blupf90 실행
blupf90 실행 화면
blupf90 실행 로그
renf90.par
BLUPF90 ver. 1.68
Parameter file: renf90.par
Data file: renf90.dat
Number of Traits 1
Number of Effects 11
Position of Observations 1
Position of Weight (1) 2
Value of Missing Trait/Observation 0
EFFECTS
# type position (2) levels [positions for nested]
1 cross-classified 13 1
2 covariable 3 1 14
3 covariable 4 1 14
4 covariable 5 1 14
5 covariable 6 1 14
6 covariable 7 1 14
7 covariable 8 1 14
8 covariable 9 1 14
9 covariable 10 1 14
10 covariable 11 1 14
11 covariable 12 1 14
Residual (co)variance Matrix
245.00
correlated random effects 2 3 4 5 6 7 8 9 10 11
Type of Random Effect: diagonal
trait effect (CO)VARIANCES
1 2 9.960 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
1 3 0.000 9.960 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
1 4 0.000 0.000 9.960 0.000 0.000 0.000 0.000 0.000 0.000 0.000
1 5 0.000 0.000 0.000 9.960 0.000 0.000 0.000 0.000 0.000 0.000
1 6 0.000 0.000 0.000 0.000 9.960 0.000 0.000 0.000 0.000 0.000
1 7 0.000 0.000 0.000 0.000 0.000 9.960 0.000 0.000 0.000 0.000
1 8 0.000 0.000 0.000 0.000 0.000 0.000 9.960 0.000 0.000 0.000
1 9 0.000 0.000 0.000 0.000 0.000 0.000 0.000 9.960 0.000 0.000
1 10 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 9.960 0.000
1 11 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 9.960
REMARKS
(1) Weight position 0 means no weights utilized
(2) Effect positions of 0 for some effects and traits means that such
effects are missing for specified traits
* The limited number of OpenMP threads = 4
* solving method (default=PCG):FSPAK
Data record length = 14
# equations = 11
G
9.9600 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000
0.0000 9.9600 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000
0.0000 0.0000 9.9600 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000
0.0000 0.0000 0.0000 9.9600 0.0000 0.0000 0.0000 0.0000 0.0000
0.0000
0.0000 0.0000 0.0000 0.0000 9.9600 0.0000 0.0000 0.0000 0.0000
0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 9.9600 0.0000 0.0000 0.0000
0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 9.9600 0.0000 0.0000
0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 9.9600 0.0000
0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 9.9600
0.0000
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
9.9600
read 8 records in 0.1718750 s, 66
nonzeroes
finished peds in 0.1718750 s, 66 nonzeroes
left hand side
0.0003 -0.0001 0.0001 0.0001 0.0001 -0.0001 -0.0000 -0.0000 0.0000 0.0002 0.0001
-0.0001 0.1005 -0.0000 -0.0001 -0.0001 0.0000 0.0000 0.0000 -0.0000 -0.0001 0.0000
0.0001 -0.0000 0.1005 -0.0000 0.0000 -0.0000 -0.0001 -0.0000 0.0000 0.0000 0.0001
0.0001 -0.0001 -0.0000 0.1006 0.0000 -0.0000 -0.0001 -0.0000 0.0000 0.0001 0.0000
0.0001 -0.0001 0.0000 0.0000 0.1006 -0.0000 -0.0001 -0.0000 0.0000 0.0000 0.0001
-0.0001 0.0000 -0.0000 -0.0000 -0.0000 0.1004 -0.0000 0.0000 -0.0000 -0.0000 0.0000
-0.0000 0.0000 -0.0001 -0.0001 -0.0001 -0.0000 0.1006 0.0000 -0.0000 0.0001 -0.0002
-0.0000 0.0000 -0.0000 -0.0000 -0.0000 0.0000 0.0000 0.1004 -0.0000 -0.0000 -0.0000
0.0000 -0.0000 0.0000 0.0000 0.0000 -0.0000 -0.0000 -0.0000 0.1004 0.0000 0.0000
0.0002 -0.0001 0.0000 0.0001 0.0000 -0.0000 0.0001 -0.0000 0.0000 0.1006 -0.0001
0.0001 0.0000 0.0001 0.0000 0.0001 0.0000 -0.0002 -0.0000 0.0000 -0.0001 0.1006
right hand side:
0.00 -0.00 0.00 0.00 0.00 -0.00 -0.00 -0.00 0.00 0.00
0.00
solution:
9.12 0.00 -0.00 0.00 -0.00 0.00 0.00 0.00 -0.00 0.00
-0.00
solutions stored in file: "solutions"
blupf90 실행 결과 : solutions
trait/effect level solution
1 1 1 9.12441059
1 2 1 0.00004354
1 3 1 -0.00440133
1 4 1 0.00439877
1 5 1 -0.00104827
1 6 1 0.00048476
1 7 1 0.00229456
1 8 1 0.00000000
1 9 1 -0.00000000
1 10 1 0.00179833
1 11 1 -0.00125139
실행 결과가 책과 다른데 이것은 책에서 EDC의 역수가 EDC를 가중치로 주어 방정식을 푼 것으로 보인다. 이 결과를 이용한 DGV 계산은 생략한다.